INDEX

Explanations

This neuron fires on imperative action words that name CRUD or API operations (e.g. create, get, insert, update, delete, select).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 radiative

0.57

STACK

0.56

 Forschungs

0.55

 मस्तिष्क

0.54

 RADIATION

0.54

 विकिरण

0.54

糁

0.54

 Seph

0.53

仿佛

0.52

radiative

0.52

POSITIVE LOGITS

 edit

1.27

 delete

1.20

findById

1.16

 Delete

1.10

delete

1.09

 CRUD

1.06

edit

1.05

ById

1.05

 deleted

1.03

删除

1.03

Activations Density 0.012%